An Integrated Approach to Home Security and Safety Systems 1st Edition

This book provides an integrated solution for security and safety in the home, covering both assistance in health monitoring and safety from strangers/intruders who want to enter the home with harmful intentions. It defines a system whereby recognition of a person/stranger at the door is done using three modules: Face Recognition, Voice Recognition and Similarity Index. These three modules are taken together to provide a percentage likelihood that the individual is in the "known" or "unknown" category. The system can also continuously monitor the health parameters of a vulnerable person living alone at home and aid them in calling for help in an emergency.

The authors have analyzed a number of existing biometric techniques to provide security for an individual living alone at home. These biometric techniques have been tested using MATLAB® image processing and signal processing toolboxes, and results have been calculated on the basis of recognition rate. A major contribution in providing security is a hybrid algorithm proposed by the author named PICA, which combines features of both PCA (Principle Component Analysis) and ICA (Independent Component Analysis) algorithms. This hybrid approach gives better performance recognition than either system alone. The second proposed hybrid algorithm for voice recognition is named as a MFRASTA algorithm by combining features of MFCC (Mel Frequency Cepstral Coefficient) and RASTA-PLP (RelAtive SpecTrA-Perceptual Linear Prediction) algorithm. After performing experiments, results are collected on the basis of recognition rate.

The authors have also proposed a third technique named as a Similarity Index to provide trust-based security for an individual. This technique is text independent in which a person is recognized by pronunciation, frequency, tone, pitch, etc., irrespective of the content spoken by the person. By combining these three techniques, a high recognition rate is provided to the person at the door and high security to the individual living independently at home. In the final contribution, the authors have proposed a fingertip-based application for health monitoring by using the concept of sensors. This application is developed using iPhone 6’s camera. When a person puts their fingertip on a camera lens, with the help of brightness of the skin, the person’s heartbeat will be monitored. This is possible even with a low-quality camera. In case of any emergency, text messages will be sent to the family members of the individual living alone by using 3G Dongle and MATLAB tool. Results show that the proposed work outperforms all the existing techniques used in face recognition, voice recognition, and health monitoring alone.

Dr. Sonali Goyal is PhD in Computer Science & Engg. She has teaching experience of 8 years and is currently working as an Assistant Professor in the computer science department at MMDU. She has published more than 10 papers in national and international journals of repute. She has published two patents which are to be awarded soon. Her areas of interest in computer science are Pervasive Computing, Bio-metrics and Data Mining.

Dr. Neera Batra has a Phd in Computer Applications. She has a decade of teaching and research experience and is currently working as an Associate Professor in computer science department at MMDU. She has published more than 40 papers in national and international journals of repute. She has published six patents which are to be awarded soon. Her areas of interest in computer science are Pervasive computing, Internet of Things, Distributed Database and Security.